Forum Discussion
NimbostratusOther methods than spanning tree to block loops in F5?
According to http://support.f5.com/kb/en-us/products/big-ip_ltm/manuals/product/tmos_management_guide_10_1/tmos_stp.html1006923 F5 supports today the following spanning tree standards:
Protocol Name IEEE Specification
Spanning Tree Protocol (STP) 802.1D-1998
Rapid Spanning Tree Protocol (RSTP) 802.1w, 802.1t, and 802.1D-2004
Multiple Spanning Tree Protocol (MSTP) 802.1s
Where MSTP is the one to choose if you for example want something similar to PVST (Per VLAN Spanning Tree) - that is over a physical link the spanning tree will block forwarding for one VLAN while it allows forwarding for another.
But in a F5-only environment (at least for the links in question) - are there other, proprietary or such, anti-loop protocols one can use in F5?
Specifically something that is subsecond in its failover time compared to the rapid settings which are (if Im not mistaken):
hellotime: 2 seconds
forwarddelay: 4 seconds
maxage: 6 seconds
For example Allied Telesis uses something they call "Protection Switching Rings" which (they claim) has a sub-50ms failover time - does F5 have something similar?
2 Replies
What_Lies_Bene1Cirrostratus
I'm afraid not. It's obviously in F5's interests to be open and support as many standards as possible to ensure their devices can integrate well in most environments, however, as technology such as your example is proprietary it's not really feasible. You wouldn't expect the F5 to be too central to the layer two infrastructure anyway and there are many ways to design and implement them to keep them 'out of the loop'.
In my experience anyway rapid is just that and tends to be faster to converge than the timers might suggest.
mikand_61525ITU-T now has this standardized and its called "Ethernet Ring Protection Switching", see link for more information:
http://en.wikipedia.org/wiki/Ethernet_Ring_Protection_Switching
So anyone with intel on when this will surface in F5 (or perhaps it already has)? :-)
